Higher Health Service Use, Mortality Rates Found Among Homebound Medicare Advantage Beneficiaries

In a national Medicare Advantage (MA) plan, homebound beneficiaries were independently associated with increased health service use and mortality, according to a study published today in Annals of Internal Medicine.¹

Past research determined that about 2 million beneficiaries are either completely or mostly homebound, while another 5.5 million need assistance or have difficulty leaving their homes.² Presumably because of restrictions related to COVID-19, the overall homebound population doubled in 2020, with greater increases affecting minority populations.³

Homebound older adults typically have several chronic conditions, functional impairment, and limited social capital.¹ Their homebound status increases their independent death risk, with nearly half of community-dwelling older adults becoming homebound in the year before death.⁴ Therefore, homebound older adults utilize care facilities more often, with health care spending more than twice that of nonhomebound older adults.¹

Currently, MA plans provide coverage to over 50% of Medicare beneficiaries, and coverage is expected to increase to approximately 60% by 2030.⁵ Consequently, MA interest in home-based care has been concurrently growing, especially among integrated payer-provider organizations.⁶ Therefore, the researchers conducted a study to determine the characteristics, predictors, prevalence, health service use, and mortality outcomes of homebound older adult beneficiaries of a large national MA plan.¹

Homebound beneficiaries of a national Medicare Advantage (MA) plan had higher health service use and mortality rates. | Image Credit: didesign - stock.adobe.com

To conduct their study, the researchers identified individuals enrolled in MA plans offered by Humana, a large national insurer. Because of their focus on homebound status, they limited the sample to beneficiaries who completed an in-home health and well-belling assessment (IHWA) between January and December 2022; eligible individuals also had to be continuously enrolled in an MA plan from January 1, 2021, through December 31, 2021.

The IHWA assessments were conducted over 45 to 60 minutes by licensed clinicians in all states except Alaska. Homebound status was determined through the self-reported frequency with which people left their homes, whether help was required, and whether they found it difficult to leave their homes in the past month.

Based on their responses, the researchers categorized the study population into 1 of 3 categories: homebound, semihomebound, or nonhomebound. Homebound people were those who rarely left their homes. Although they left their homes more often, semihomebound people were at risk of becoming homebound since they needed assistance or considered it difficult. The remaining people were considered nonhomebound.

Additionally, the researchers compared demographic and clinical characteristics, health service use, and mortality outcomes by homebound status; they obtained sociodemographic, clinical, and health care use data from Humana’s internal claims and membership data. For each patient, the researchers computed the CMS Hierarchical Condition Category (HCC) score and determined their frailty level based on a claims-based frailty index. Based on their frailty score, they categorized the beneficiaries into one of these categories: nonfrail (< 0.15), prefrail (0.15-0.24), mildly frail (0.25-0.34), and moderately to severely frail (≥0.35).

Of 5,722,966 active MA beneficiaries in 2022, 1,456,019 were not targeted to receive an IHWA. After further exclusions, 2,435,519 beneficiaries were targeted for an IHWA and therefore eligible for the study (42.6% of all active MA beneficiaries in 2022). The researchers noted that the IHWA was completed by 514,188 people, which accounted for 21.1% of targeted beneficiaries and 9.0% of all active MA beneficiaries. Of those who completed the IHWA, 401,136 (78.0%) were not homebound, 69,740 (13.6%) were semihomebound, and 43,312 (8.4%) were homebound.

Compared to nonhomebound beneficiaries, those defined as homebound or semihomebound were more likely to be female (67.2% vs 61.8% vs 51.9%), older (74.4 vs 73.3 vs 71.4 years), and either low-income or dual-eligible for Medicare and Medicaid (44.8% vs 42.4% vs 31.4%). Additionally, homebound or semihomebound beneficiaries had higher weighted HCC scores (6.01 vs 5.71 vs 3.84) and a higher prevalence of chronic conditions; dementia was about 4 times more prevalent among homebound adults. As for frailty, homebound and semihomebound beneficiaries were more likely to be mildly (20.3% vs 17.4% vs 6.0%) or moderately to severely (5.8% vs 3.5% vs 0.5%) frail.

Therefore, multiple factors were independently associated with being homebound, including female sex (OR, 1.36; 95% CI, 1.35-1.37), dementia (OR, 2.36; 95% CI, 2.33-2.39), low-income status or dual-eligibility (OR, 1.56; 95% CI, 1.55-1.57), and moderate to severe frailty (OR, 4.32; 95% CI, 4.19-4.45). Similarly, in a multivariable logistic regression, the researchers determined that homebound status was associated with increased odds of any emergency department visit (OR, 1.14; 95% CI, 1.14-1.15), inpatient hospital admission (OR, 1.44; 95% CI, 1.42-1.46), or skilled-nursing facility admission (OR, 2.18; 95% CI, 2.13-2.23). Homebound status was also associated with the highest odds of death (OR, 2.55; 95% CI, 2.52-2.58).

The researchers explained that their findings have policy implications. They noted that IHWA data on functional status could be incorporated into payment and risk adjustment models “to support more accurate payment and the more intensive care models needed to meet the needs of homebound adults.”

Conversely, the researchers acknowledged their limitations, one being that they used data from a single national MA plan. Due to sampling issues, the findings may not be generalizable to the full plan population or the broader MA population. Despite their limitations, the researchers made suggestions for future actions.

“As MA becomes the majority payer source for Medicare beneficiaries, attention to providing appropriate care delivery to this high-need, high-cost population is warranted,” the authors concluded.

References

1. Leff B, Ritchie C, Szanton S, et al. Epidemiology of homebound population among beneficiaries of a large national Medicare Advantage plan. Ann Intern Med. Published online August 12, 2024. doi:10.7326/M24-0011
2. Ornstein KA, Leff B, Covinsky KE, et al. Epidemiology of the homebound population in the United States. JAMA Intern Med. 2015;175(7):1180-1186. doi:10.1001/jamainternmed.2015.1849
3. Ankuda CK, Leff B, Ritchie CS, Siu AL, Ornstein KA. Association of the COVID-19 pandemic with the prevalence of homebound older adults in the United States, 2011-2020. JAMA Intern Med. 2021;181(12):1658-1660. doi:10.1001/jamainternmed.2021.4456
4. Soones T, Federman A, Leff B, Siu AL, Ornstein K. Two-year mortality in homebound older adults: an analysis of the National Health and Aging Trends Study. J Am Geriatr Soc. 2017;65(1):123-129. doi:10.1111/jgs.14467
5. Ochieng N, Fuglesten Biniek J, Freed M, et al. Medicare Advantage in 2023: enrollment update and key trends. KFF. August 9, 2023. Accessed July 11, 2024. www.kff.org/medicare/issue-brief/medicare-advantage-in-2024-enrollment-update-and-key-trends/
6. Volpp KG, Diamond SM, Shrank WH. Innovation in home care: time for a new payment model. JAMA. 2020;323(24):2474-2475. doi:10.1001/jama.2020.1036